James Cameron, after his blockbuster movie Titanic, has spent 15 years working on a $430M Avatar ... in 3D no less.

So I figure I should larn something 'bout this technology, eh? First off, we got us two eyes and each eye sees something different. Similar, but different.             >That's why we see 3 dimensional, right? Right. So cameras which record 3D movies gotta have two eyes as well ... like this See? They got a right and a left "eye". Now, the problem is to get each separate image to the appropriate eye, when you watch the movie.     To do this, each camera "eye" records a polarized version of the scene. A vertical polarization and a horizontal polarization - one for each of your eyes. >Huh?

I'm sure you know about polarized sunglasses. The glare off the surface of a lake, for example, is partially polarized. Horizontally polarized light (for example) won't pass through glasses that have a vertical polarizer ... and the glare is (partially) horizontally polarized as a result of the reflection from the surface. So you just buy glasses that eliminate the polarization associated with the glare.

Figure 1

>So, to watch 3D movies I buy sunglasses? No. Both lenses in polarized sunglasses have the same polarization direction. You'll need special glasses where each lens has a different polarization ... that matches that of the 3D camera's "eyes". >I assume you'll be watching that 3D movie, eh? You betcha!

I should point out that the movie is mostly computer generated images ... so the creators had to generate a "double image", one for each eye.
I might also point out that Avatar can also be seen in iMax theatres.
Did y'all know that iMax, a Canadian invention, was first seen at the 1967 Montreal Expo?

more on Polarizers

A polarizer contains molecules, aligned in a particular direction, which absorb (or reflect) light that has its direction of oscillation in the same direction as the molecular alignment.
Some polarizers contain very fine wires which will absorb light oscillationg in their direction. Others contain silver nanoparticles embedded in the polarizer film.
Some crystal structures interact with incoming light differently, depending upon the direction of oscillation of the light, so that ...

>And those 3D glasses? What about them?
Uh ... wait till I get to the movie theatre, eh?

In any case, one might expect that, when the incoming light has light in all directions, picking out just the vertically oriented waves would greatly reduce the intensity of the light that gets through the polarizer. However, even light waves that oscillate at an angle to the polarizer direction will have some component in that direction
... and that component will pass through the polarizer.

>Huh? All light wave oscillations will have some component in the polarizer direction ... ... except for perpendicular oscillations. Those components do get through the polarizer ... though reduced in intensity. The amplitude of that component varies with the angle of oscillation. The reduction in amplitude is cos(θ) and the intensity is proportional to the square of the component and the average of cos2(θ) is 1/2.

>So the transmitted light has 50% the intensity of the incident light, right?
Uh ... yes, in a perfect world.

See Figure 1, above? All components in the direction of the grey lines are absorbed by the polarizer ... and only the perpendicular oscillations get through.

I also discover that, in some polarizer materials, light travels at different speeds ... depending upon the direction of polarization.
For example, verticall oscillations will travel at one speed and horizontal oscillations at a slower speed.
In that way, one can arrange the thickness of the polarizer so that the slower light wave is displaced by a quarter wavelength.
These so-called quarter wave plates can be used to change linearly polarized light to "circularly" polarized light.

>I have no idea what that means!

• Light is an electromagnetic wave.
• There's an electric vector associated with the wave.
• This electric vector oscillates in a particular direction.
• This direction of oscillation is what we are referring to when, above, we talked about the "direction of oscillation".
• When light (with a host of electric vectors, in all directions) passes through a linear polarizer, the emerging light has its electric vector oscillating in just one direction.
  (The component of Electric Vector in the direction of the polarizer "lines" are absorbed or reflected.
  That's why, in Fig. 1, the direction of oscillation of the emerging light is perpendicular to the polarizer "lines".)
• If that linearly polarized light passes through a quarter wave plate, we can get the electric vector to oscillate in a circular manner, like so:
  (That's the result of adding two waves displaced by a quarter wavelength.) Direction(s) of Oscillation of the Electric Vector

3D TV

I understand they're working on 3D TV, where each eye has a slightly different view of what's on the screen.
In front of the TV screen is a sheet with a jillion tiny lenses which project the scene in slightly different directions.

>So each eye sees the same thing?
Apparently, the pixels which make up the scene come in pairs.
There are alternate columns of pixels: even-numbered columns display the left-eye images and odd-numbered the right-eye images.
Magically (!), the lenses project each image in a slighlty different direction so your left eye sees ...

>Sees just the left-eye image, right?
Apparently. Early stereoscopic lens projectors provided just two images / directions; one for each eye.
However, the viewer had to sit directly in front of the TV. More recent stuff has up to sixteen projected images.
However, you may still have to sit at a particular location to get the optimal 3D effect.

>No glasses required?
Apparently.

>And does it work?
Apparently ... actually, it seems like magic.